Predictors of Use of Highly Active Antiretroviral Therapy (HAART) Among Persons with AIDS in San Francisco, 1996-1999

Highly active antiretroviral therapy (HAART) has been shown to reduce viral load and increase CD4 lymphocytes in persons infected with HIV, delay onset of AIDS and prolong survival with AIDS. In San Francisco, as well as elsewhere, AIDS incidence and mortality have declined substantially since 1996 due to HAART.

Yet, many other variables influence whether HAART is used, including access to care, physician experience, and patient characteristics. This study used the population-based AIDS registry in San Francisco to evaluate the prevalence and predictors of HAART use both before and after an AIDS diagnosis.

Active surveillance of AIDS was conducted by Department of Public Health staff who reviewed pathology and laboratory reports at ten acute hospitals, eight community free-standing clinics and ten private medical practices. Passive surveillance included reporting of AIDS by providers at the remaining health facilities. AIDS cases were also identified through review of death certificates, reports from other health departments, and retrospective evaluations or special studies. Information regarding use of HAART and CD4 cell counts was collected through review of medical records and laboratory reports at the time of the initial AIDS case report and periodically thereafter. The type and start date of each HAART regimen was collected at treatment facilities where the diagnosis was made or where patients received care.

Adult and adolescent AIDS patients age 13 or older who were diagnosed between 1996 and 1999 or who were diagnosed before 1996 and survived beyond January 1, 1996 were included in the analysis. Initiation of HAART before diagnosis of AIDS was assessed among persons diagnosed with AIDS between 1996 and 1999. Initiation of HAART after AIDS diagnosis was evaluated among persons living with AIDS between 1996 and 1999 who started HAART after AIDS or who never received HAART.

In total, 8,410 AIDS patients were known to be alive between January 1, 1996 and December 31, 1999. Of these, 94 percent were male, 68 percent were white and 72 percent were men who had sex with men (MSM). Almost half of these cases were diagnosed in private practice. Nearly one-third were uninsured at the time of diagnosis. Of the 6,467 persons living with AIDS in 1996, 2,658 (41 percent) were reported to have used HAART (before, on or after AIDS diagnosis). In 1997, 3,894 (61 percent) of the living AIDS patients had used HAART. The prevalence of HAART use increased again in 1998 to 4,517 (68 percent), and to 4,847 (72 percent) in 1999.

Specific data of importance on the use of HAART before diagnosis of AIDS included the fact that only 5 percent of patients diagnosed in 1996 had initiated HAART, as compared with 26 percent in 1999. Regression analysis revealed, African-Americans (odds ratio (OR), 0.59, 95% CI); injection drug users (heterosexual injection drug user: OR, 0.59; 95%CI; MSM injection drug user: OR, 0.66 95% CI, respectively), and those without insurance at time of diagnosis (OR, 0.61; 95% CI) were less likely to use HAART before AIDS diagnosis. There were differing rates among private providers and public providers, with those using private physicians, county hospitals and community clinics more likely to initiate HAART before AIDS than those using private hospitals.

Between 1996 and 1999, there were 4,627 persons with AIDS who initiated HAART after AIDS and 2,940 persons who did not receive HAART. In the proportional hazards analysis, African-Americans were the most likely not to receive HARRT, followed by injection drug users, homeless persons, those with public insurance and those with higher CD4 counts. HAART was initiated sooner after AIDS among Latinos; persons aged 40-49, and those diagnosed at private physician's offices.

The authors offer a lengthy discussion of differentials of treatment with HAART among various groups in San Francisco. They conclude that, although HAART use in San Francisco increased rapidly during the study period, not all groups benefited equally from the availability of treatment. They recommend that "increasing use of antiretroviral therapy among disenfranchised populations should be implemented including case management and adherence programs."

Back to other CDC news for November 20, 2001

Previous Updates | Search the CDC archive